Stratos Ioannidis

Fuzzy supervisory control of manufacturing systems

A supervisory controller is derived for scheduling (single/multiple-part-type, reentrant) production networks. The supervisory controller is used to tune a set of lower level distributed fuzzy control modules that reduce work-in-process (WIP) and synchronize the production systems operation. The overall production-control system is viewed as a two-level surplus-based system with the overall control objective to keep the WIP and cycle time as low as possible, while maintaining quality of service by keeping backlog to acceptable levels. The production rate in each production stage is controlled to satisfy demand, avoid overloading, and eliminate machine starvation or blocking. The systems improvement is demonstrated using a set of performance measures. Extensive simulation results show that the supervisory controller, when compared with the single-level distributed fuzzy controllers reduces WIP and cycle time while keeping backlog to acceptable levels.

Fuzzy work-in-process inventory control of unreliable manufacturing systems

Abstract We consider single and multiple part type production lines and networks with finite buffers and unreliable machines. Three fuzzy control modules, namely, line , assembly , and disassembly controller, are developed. The objective is to keep the work-in-process (WIP) inventory and cycle time at low levels, along with high machine utilization and throughput. This is achieved by adjusting the processing rate of each production stage so that workflow is balanced and the extreme events of machine starving or blocking are reduced. The approach is extensively tested via simulation. After a series of simulation runs, it has been observed that the proposed approach outranks other control policies in keeping the WIP inventory low.

Analysis of Admission and Inventory Control Policies for Production Networks

Problems of inventory control and customer admission control are considered for a manufacturing system that produces one product to meet random demand. Four admission policies are investigated: lost sales, complete backordering, randomized admission, and partial backordering. These policies are combined with an integral inventory control policy, which releases raw items only when an incoming order is accepted and keeps the inventory position (total inventory minus outstanding orders) constant. The objective is to determine the inventory level and the maximum number of backorders, as well as the admission probability that maximize the mean profit rate of the system. The system is modeled as a closed queueing network and its performance is computed analytically. The optimal parameters for each policy are found using exhaustive search and convex analysis. Numerical results show that managing inventory levels and sales jointly through partial backordering achieves higher profit than other control policies.

Revenue management in single-stage CONWIP production systems

We study a joint admission/inventory control problem for a manufacturing system producing one product to meet random demand. The system employs a constant work- in-process policy (CONWIP) whereby the total inventory of raw material and finished items is kept constant, and accepts orders only as long as the backlog is below a certain level. The objective is to determine the CONWIP and backlog levels that maximize the mean profit rate of the system. The system is modelled as a single server with a finite queue. It turns out that the mean profit rate is either concave or decreasing in one control parameter and also decreasing for large values of the other control parameter. A simple algorithm is developed which tracks down the globally optimum design in finite time. Numerical results show that the joint admission/inventory control policy achieves higher profit than other production control policies that have been examined in the literature.

Coordinating quality, production and sales in manufacturing systems

We explore the benefits of jointly designing quality tolerances, customer admission and production control policies in manufacturing systems producing a single product to meet demand. These problems have been addressed separately in the past. We consider a simple admission/production control policy whereby the system produces until stock reaches a certain level and accepts orders until the backlog reaches another critical level. We model the system using queueing theory and propose an easily implementable procedure for selecting the optimal quality tolerances and the critical stock and backlog levels. From theoretical and numerical results, it appears that the proposed policy achieves a higher profit than other manufacturing practices, in which there is little or no coordination between the production, sales and quality control departments.

Control Policies for Single-Stage Production Systems with Perishable Inventory and Customer Impatience

We consider problems of inventory and admission control for make-to-stock production systems with perishable inventory and impatient customers. Customers may balk upon arrival (refuse to place orders) and renege while waiting (withdraw delayed orders) during stockouts. Item lifetimes and customer patience times are random variables with general distributions. Processing, setup, and customer inter-arrival times are however assumed to be exponential random variables. In particular, the paper studies two models. In the first model, the system suspends its production when its stock reaches a safety level and can resume later without incurring any setup delay or cost. In the second model, the system incurs setup delays and setup costs; during stockouts, all arriving customers are informed about anticipated delays and either balk or place their orders but cannot withdraw them later. Using results from the queueing literature, we derive expressions for the system steady-state probabilities and performance measures, such as profit from sales and costs of inventory, setups, and delays in filling customer orders. We use these expressions to find optimal inventory and admission policies, and investigate the impact of product lifetimes and customer patience times on system performance.

Joint production and quality control in production systems with two customer classes and lost sales

This article considers a single-product, make-to-stock manufacturing system facing random demand from two customer classes with different quality cost and profit parameters. Each outgoing product is inspected and graded on the basis of quality. Each customer class can only be served from the inventory of products of certain quality grades, if any exist, and the system may reserve a fraction of the inventory of some quality grade for customers of a certain class. The problem is one of finding a product quality grading plan and production and inventory rationing policies to maximize the mean profit rate of the system. The structures of the optimal production and order admission control policy are investigated numerically for the case when a specific quality grading plan is used. Based on this investigation, some simple and efficient threshold-type control policies are proposed. From numerical results, it appears that the proposed approach of coordinated quality, production control, and inventory control achieves higher profit than other manufacturing practices, in which there is little or no coordination between the production and quality control departments.

Work-in-process scheduling by evolutionary tuned distributed fuzzy controllers

This paper presents an evolutionary algorithm (EA) strategy for the optimization of generic work-in-process (WIP) scheduling fuzzy controllers. The EA strategy is used to tune a set of distributed fuzzy control modules that are responsible for WIP scheduling. The objective is to control the production rate in each processing stage in a way that satisfies the demand for final products, while reducing WIP within the production system. The EA identifies the parameters for which a set of fuzzy controllers performs optimal with respect to WIP and backlog minimization. The proposed EA strategy is compared to heuristically tuned distributed fuzzy controllers. Extensive simulation testing shows that the EA strategy significantly improves the systems performance

The Effects of Quality on Market Share and Profitability in Single Stage Make-to-Stock Production Systems

We develop simple models for understanding how the dynamics of quality may affect customer satisfaction and profitability in make-to-stock manufacturing systems. We study a Markovian, single stage system facing random demand. Any demand not satisfied immediately from stock is lost to competitors. The market is assumed to be finite and comprises both regular and occasional customers. Regular customers have higher mean demand rates than occasional customers. Each outgoing product is inspected and classified as high quality, medium quality, or nonconforming. The customer who purchases an item joins the regular or the occasional class, with corresponding probabilities which depend on the quality level and on past customer state. The higher the quality level, the higher the probability for a customer to remain or become a regular customer. Our goal is to investigate the structure of the optimal production, order satisfaction and quality control policy in order to maximize the average profit rate of the system. We investigate numerically the structure of the optimal policy using stochastic dynamic programming.

Inventory and order admission control in manufacturing systems with two customer classes and setup times

In this paper we examine a Markovian singlestage manufacturing system, with setup times producing a single item to satisfy demand of two different customer classes. We investigate numerically the structure of the optimal inventory and order admission control policy. Due to the complexity of the optimal policy a simple heuristic policy is proposed. Numerical results show that the proposed policy can be a very good approximation of the optimal policy.